SAT0053 ESTIMATING REAL-WORLD UNMET NEEDS FOR REACHING REMISSION IN THE FIRST YEAR FOLLOWING EARLY RA DIAGNOSIS: RESULTS FROM THE CANADIAN EARLY ARTHRITIS COHORT (CATCH)

Abstract

Background:Several composite RA disease activity indices are commonly used in clinical practice and research. Different disease activity indices however can be inconsistent in classifying remission (REM).Objectives:1) Compare remission prevalence across 4 common RA indices; 2) compare changes in remission across indices; and, 3) Identify predictors of persistent active disease across all indices, in real-world early RA patients over 1 year follow up.Methods:Data were from patients with early RA (symptoms < 1 year) enrolled in the Canadian Early Arthritis Cohort (CATCH) between 2007 and 2018. Participants had active disease at enrolment, were treated with csDMARDs and completed standardized clinical assessments every 3-months. Remission status was assessed using 4 indices: 1) DAS28< 2.6 OR DAS28CRP < 2.5, 2) CDAI ≤ 2.8, 3) SDAI≤ 3.3, and 4) ACR/EULAR Boolean remission – SJC28, TJC28, CRP, PGA all ≦1. T-tests/ chi-squared tests were used to compare differences in remission prevalence by 1 year, and changes in remission before and after a QI program. Logistic regression was used to identify predictors of persistent active disease on all 4 indices.Results:1202 adults were eligible for this analysis. At enrolment, 877 (73%) were women, mean (sd) age was 55 (14), average disease activity was high (DAS28 5.1 (1.4); CDAI 27 (14); SDAI 29 (15)). Prevalence of remission by 12-months follow up was 14-21% higher when estimated with the DAS28 compared with CDAI, SDAI and Boolean criteria, and 378 (31%) did not achieve remission according to any of the 4 indices (Fig 1). Improvement in remission after a QI program however was similar across all 4 indices(~+15-17%). In adjusted logistic regression, Persistent active disease across all measures was most strongly associated with positive serostatus and smoking in men, and with obesity and more tender joints in women. Pain and lower education were predictors in BOTH men and women (Table 2)Table 1.Multivariable Logistic Regression Predicting Persistent Active Disease by 12-months across ALL RA indicesConclusion:In the absence of a single “best measure” that also takes in to account the patient’s perspective, we estimate unmet needs for achieving remission in the first year of follow up in 1 in 3 ERA patients who did not achieve remission by ANY of the 4 indices.